A majority of deaf students leave high school reading at or below a fourth grade level, but some deaf individuals do become highly proficient readers. There is disagreement about the causes of this reading difficulty, and by association, disagreement about the effectiveness of different strategies for teaching reading to deaf children. One hypothesis is that reading difficulties in deaf individuals are caused by a lack of direct access to sound information of the language being read. A second hypothesis is that reading difficulties arise because many deaf children are not proficient in any language at the point in time when they begin to learn to read, whether it be a signed or spoken language. Decades of research have found varying degrees of support for both of these hypotheses, but conflicting results prevent consensus. Given that hearing individuals can attain reading proficiency or deficiency in a variety of ways, it is possible that different deaf individuals, especially those from different language backgrounds, may read successfully using different mechanisms. Prior research into the ways that deaf people read has primarily used behavioral tasks to measure reading processes. However, behavioral tasks do not directly assess the real-time neurocognitive processes underlying reading. Event-related potentials (ERPs), recorded while a subject reads, provide a direct millisecond-by- millisecond record of the brain's electrica activity. ERPs respond differently to specific aspects of language, including sound (phonology), spelling (orthography), meaning, and grammar, and the size of an ERP response is a function of a person's skill in and exposure to a language. We will use ERPs, recorded while deaf and hearing subjects read English words and sentences, to measure subjects' real-time sensitivity to phonological and orthographic information about words (via priming responses to words related in phonology or orthography), as well as to the concepts of meaning and grammar (via errors of meaning and grammar in sentences). We will determine which online language processing mechanisms, as measured by ERPs, are the best predictors of standardized reading comprehension scores. Specifically, we will address two questions: whether deaf and hearing individuals reach skilled reading proficiency by using similar or different online language processing mechanisms, and whether deaf individuals from different language backgrounds (i.e., from signed versus spoken language backgrounds) reach skilled reading proficiency by different routes. This study will provide a clearer picture of the mechanisms by which deaf people read, how these mechanisms relate to reading proficiency and how a deaf person's language background contributes to the neural processing of reading. To develop more effective methods for teaching deaf children to read, we must first understand how some deaf people read more successfully than others, and the proposed work will do so in a novel way.